期刊文献+

高性能低功耗高速缓存的V-LRU RAM单周期清零技术

Clearing V-LRU RAM in a Single Clock Cycle for High-Performance Low-Power Cache
原文传递
导出
摘要 提出并实现了一种高速缓存的V-LRU RAM单周期清零技术。运行操作系统的CPU在不同任务之间切换时,需要对V-LRU RAM清零。使用传统的计数器依次清空V-LRU RAM的各行,CPU会白白浪费很多个时钟周期。在一个时钟周期对V-LRU RAM清空,可以大大提高CPU的性能。在四路组相联的高速缓存设计中,容量为16k、8k和4k字节时,使用该技术可以将以前的256、128和64个时钟周期降低到只有1个时钟周期。基于SMIC 0.13μm工艺,实现该技术的硬件电路面积为6 312.8μm2,且高速缓存的缺失率保持在非常低的水平。这种技术同样适用于对RAM需要单周期清空的场合。 A technique to clear V-LRU RAM for high performance cache in a single clock cycle was proposed and implemented.The CPU running operating system need to clear V-LRU RAM when switching between different tasks.When traditional counter is used to clear every line of V-LRU RAM one by one,a large number of clock cycles for CPU was wasted.Clearing V-LRU RAM in one clock cycle could significantly improve performance of the CPU.For a 4-way set-associative cache with a capacity of 16-kByte,8-kByte,and 4-kByte,using the proposed technique,the number of clock cycles could be reduced from 256,128 and 64 to only one.The hardware circuit to implement the technology occupied a chip area of 6312.8 μm2 in SMIC's 0.13 μm process,and miss rate of the cache kept at a very low level.This technology could also be used for other applications where clearing RAM in a single cycle is needed.
出处 《微电子学》 CAS CSCD 北大核心 2012年第1期97-101,共5页 Microelectronics
关键词 单周期清零 V-LRU RAM 高速缓存 时钟周期 缺失率 Clearing in single clock cycle V-LRU RAM Cache Clock cycle Miss rate
  • 相关文献

参考文献13

  • 1BAO D X, LI X M. A cache scheme based on LRU- like algorithm [C] // Proc IEEE Int Conf Inform Autorru Harbin, China. 2010: 2055-2060.
  • 2董磊,岑俊龙,涂志娣,李可,梁松海.高压缩V-LRU算法的研究与实现[J].微电子学,2011,41(5):731-735. 被引量:1
  • 3KEDZIERSKI K, MORETO M, CAZORLA F J, et al. Adapting cache partitioning algorithms to pseudo- LRU replacement policies [C] // IEEE Int Parallel Distributed Process Syrup. Atlanta, GA, USA. 2010: 1-12.
  • 4CARLSON I, ANDERSSON S, NATARAJAN S, et al. A novel five-transistor (5T) SRAM cell for high performance cache [C]// Proc 30th Europ Sol Sta Circ Conf. Leuven, Belgium. 2004: 101-102.
  • 5GUPTA V, ANIS M. Statistical design of the 6T SRAM bit cell [J] IEEE Trans Circ Syst, 2010, 57 (1) : 93-104.
  • 6郝玉艳,彭蔓蔓.混合Cache的低功耗设计方案[J].计算机工程与应用,2009,45(20):68-70. 被引量:2
  • 7李芳,徐丽,陈亮亮.LRU近似算法的研究[J].现代电子技术,2009,32(10):36-38. 被引量:7
  • 8HENNESSY J L, PATTERSON D A. Computer architecture: a quantitative approach [M]. 4th Ed. New York.. Morgan Kaufmann Publishers, 2007.. 288- 342.
  • 9夏宏,苏林萍.Cache低功耗技术研究[J].计算机工程与应用,2005,41(23):118-120. 被引量:4
  • 10田芳芳,樊晓桠,靖朝鹏,靳战鹏.一种基于流水线的指令CACHE优化设计[J].微电子学与计算机,2006,23(1):93-96. 被引量:3

二级参考文献36

  • 1马志强,季振洲,胡铭曾.基于分类访问的低功耗联合式cache方案[J].哈尔滨工程大学学报,2007,28(1):21-25. 被引量:3
  • 2朱平,吴碧伟.磁盘缓存管理机制研究[J].计算机工程与应用,2004,40(20):47-49. 被引量:5
  • 3孟宪福.基于优先级和LRU算法的持续CORBA对象管理策略研究[J].大连理工大学学报,2005,45(6):907-911. 被引量:1
  • 4肖斌,方亮,柴亦飞,陈章龙,涂时亮.低功耗的可重构数据Cache设计[J].计算机工程与设计,2007,28(7):1508-1510. 被引量:5
  • 5Intel.Intel XScale microarchitecture.2001.
  • 6Zhang C,Vahid F,Najjar W.A highly-configurable cache architecture for embedded systems[J].ACM SIGARCH Computer Architecture News,2003,31(2): 136-146.
  • 7Albonesi D H.Selective cache ways: On-demand cache resource allocation[J].Journal of Instruction Level Parallelism,2000:248-259.
  • 8Zhang C,Vahid F,Najjar W.Energy benefits of a configurable line size cache for embedded systems[C]//IEEE International Symposium on VLSI Design, February 2003.
  • 9Malik A,Moyer B,Cermak D.A low power unified cache architecture providing power and performance flexibility[C]//Proceedings of the 2000 International Symposium on Low Power Electronics and Design, 2000: 241-243.
  • 10Smith J E,Dhodapkar A S.Dynamic microarchitecture adaptation via co-designed virtual machines[C]//2002 IEEE Solid-State Circuits Conference, Digest of Technical Papers, ISSCC, 2002:198-199.

共引文献12

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部